This paper compares rival sovereign default models that differ in how country-, region- and time-specific effects are treated. The quality of the models is gauged using inference-based criteria and the plausibility of estimates. An out-of-sample forecast evaluation framework is deployed based on...